Acta Ophthalmologica 2014
Retinal vessel oxygen saturation and vessel diameter in retinitis pigmentosa Thor Eysteinsson,1,2 Sveinn H. Hardarson,2 David Bragason2 and Einar Stefansson2 1
Physiology, University of Iceland, Reykjavik, Iceland Ophthalmology, University of Iceland/Landspitali University Hospital, Reykjavik, Iceland
2
ABSTRACT. Purpose: To assess retinal vessel oxygen saturation and retinal vessel diameter in retinitis pigmentosa. Methods: A retinal oximeter (Oxymap ehf., Reykjavik, Iceland) was used to measure retinal vessel oxygen saturation and vessel diameter in ten patients with retinitis pigmentosa (RP) (mean age 49 years, range 23–71 years). Results were compared with age- and gender-matched healthy individuals. All patients had advanced stage of the disease with visual fields restricted to the macular region. Results: Oxygen saturation in retinal venules was 58.0 6.2% in patients with RP and 53.4 4.8% in healthy subjects (p = 0.017). Oxygen saturation in retinal arterioles was not significantly different between groups (p = 0.65). The mean diameter of retinal arterioles was 8.9 1.6 pixels in patients with RP and 11.4 1.2 in healthy controls (p < 0.0001). The corresponding diameters for venules were 10.1 1.2 (RP) and 15.3 1.7 (healthy, p < 0.0001). Conclusions: Increased venous saturation and decreased retinal vessel diameter suggest decreased oxygen delivery from the retinal circulation in retinitis pigmentosa. This is probably secondary to tissue atrophy and reduced oxygen consumption. Key words: oxygen – retinal oximetry – retinitis pigmentosa – vessel diameter
Acta Ophthalmol. 2014: 92: 449–453 ª 2014 Acta Ophthalmologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
doi: 10.1111/aos.12359
Introduction Retinitis pigmentosa (RP) is a hereditary disease characterized by slow, progressive degeneration of photoreceptors, which leads to night blindness, constriction of visual fields and eventually reduction in visual acuity (Hartong et al. 2006; Ferrari et al. 2011). A large number of mutations in several of the genes expressed in retinal photoreceptors and retinal pigment epithelial cells have been shown to cause retinitis pigmentosa (Ferrari et al. 2011). Histopathologic examinations indicate that all layers of the retina are affected, with the most profound changes in the
photoreceptor outer segments, outer nuclear layer, less changes in the ganglion cell layer and probably the least in the inner nuclear layer (Santos et al. 1997; Milam et al. 1998; Humayun et al. 1999). Choroidal and retinal blood flow has been found to decrease in RP. Pulsatile ocular blood flow is reduced and this most likely reflects decreased choroidal blood flow (Langham & Kramer 1990; Schmidt et al. 2001). More recent studies have found decreased blood velocity in the ophthalmic artery and short posterior ciliary arteries (Cellini et al. 2010) and decreased subfoveal choroidal blood flow (Falsini et al. 2011).
Additionally, Zhang et al. (2013) found decreased blood flow in a combined measurement of the retina and choroid with an MRI technique; most of the decrease probably reflecting decreased choroidal blood flow. Histopathologic changes have been seen in the eyes of patients with RP, in which the choroidal capillaries are missing in areas corresponding to regions that have lost photoreceptors, and where bony spicules have formed (Henkind & Gartner 1983; Li et al. 1995; Milam et al. 1998). Recent studies with OCT have shown decreased thickness of the choroid in RP (Adhi et al. 2013; Ayton et al. 2013; Dhoot et al. 2013). Grunwald et al. (1996) measured the retinal circulation with a laser Doppler technique and found decreased volumetric blood flow in the major retinal venules in RP. Additionally, blood velocity has been found to be lower in retinal vessels (Beutelspacher et al. 2011) and the central retinal artery (Akyol et al. 1995). In theory, the blood flow changes could be either primary and causative or secondary, an adjustment to the decreased demand for blood and oxygen due to retinal atrophy and decreased function. Oxygen supply to the outer retina (photoreceptors) derives primarily from the choroid, while the inner retina is mostly supplied by the retinal vasculature (Linsenmeier & Padnick-Silver 2000). The boundaries between these two sources of oxygen can shift whether oxygen supply or demand changes in either the inner or outer retina. Reduced oxygen consumption in the photoreceptor layer causes oxygen from the choroid to
449
Acta Ophthalmologica 2014
reach the inner retina by diffusion, reducing the need for oxygen delivery from the retinal circulation. This could result in reduced oxygen delivery by the retinal circulation and would be seen in reduced arteriovenous difference in oxygen saturation and/or reduced blood flow and vessel diameters. Another process that may be involved is that the loss of photoreceptors that occurs in RP leads to reduced neuronal activity and metabolism in inner retinal neurones, and this affects oxygen delivery by the inner retinal vasculature. In this study, we use a non-invasive retinal oximetry technique (Hardarson et al. 2006; Geirsdottir et al. 2012) to measure oxygen saturation and diameter in retinal vessels in patients with RP. The results of this study were presented previously in an abstract form (Eysteinsson et al. 2012).
570 nm
600 nm
Methods Retinal vessel oxygen saturation and vessel diameter were measured in 10 patients with RP (mean age 49 years, range 23–71 years) and compared with 10 age and gender-matched healthy individuals. A diagnosis of RP was based on the clinical findings, which included typical funduscopic appearance, compromised peripheral and night vision, and severely reduced or no detectable scotopic and photopic aand b-waves of the electroretinogram. Visual fields as determined by standard Goldmann perimetry (II and V, 4e; Octopus 101, Haag-Streit AG, Koneitz, Switzerland) were in most cases severely restricted. Subjects with other conditions affecting the retina, diabetes, severe cardiovascular or respiratory disease were excluded from the study. The dual wavelength retinal oximeter (Oxymap ehf.) consists of two digital cameras (Insight IN1800, Diagnostic Instruments Inc., Sterling Heights, MI), a custom-made optical adapter, a beam splitter and two narrow band-pass filters, coupled to a fundus camera (Topcon TRC-50DX). The oximeter captures two fundus images simultaneously, at 570 and 600 nm wavelengths, as shown in Fig. 1. Specifically designed software (OXYMAP Analyzer software 2.2.1, version 3847, Oxymap ehf.) automatically selects measurement points on the images and measures brightness on
450
Fig. 1. Retinal oximetry from a patient with RP. The images in the upper panel were obtained via 570 nm (left) and 600 nm (right) filters simultaneously and were used for calculation of ODR. A colour-coded map of haemoglobin oxygen saturation is shown in the lower panel.
the measured vessels and to the side of the vessels at each wavelength. The ratio of light absorbance at the two wavelengths is approximately linearly related to oxygen saturation (Hardarson et al. 2006; Geirsdottir et al. 2012). The pupils of the subjects were dilated with 1% tropicamide (Mydriacyl; S.A. Alcon-Couvreur N.V., Puurs, Belgium), which was in some cases supplemented with 10% phenylephrine hydrochloride (AK-Dilate, Akorn Inc., Lake Forest, IL). Images were taken in a dark room, with the only lights coming from the fundus camera and computer screen. The lowest useable setting of the aiming light was used, and the flash intensity setting was set at 50 Ws. A band-pass filter with 80 nm full width at half maximum transmittance and 585 nm centre wavelength was used to limit unnecessary light exposure from aiming light or flash. The small aperture setting of the fundus camera was used. Small pupil setting was used only if needed. Images were first taken of the right eye and then the left eye. In general, five images were taken of each eye, most with different alignments. Images used for analysis were in general number two or five in the sequence of images, sometimes later (if more images had to be
taken to achieve good quality). Time between images was approximately 30 seconds. The eye with the better quality fundus image and more measureable vessels was chosen for analysis for each patient. The analysed images had the optic disc centred in the image. All major arterioles and venules, above 6 pixels diameter (approximately 56 µm), were measured, if possible. Measurements were made from close to the optic disc and up to the next vessel branching (50–200 pixels length limit). Each patient with RP was matched to one or an average of two or three healthy controls of the same age and gender. In some patients, a part of the vasculature was obscured by cataract. Fewer vessels were above the diameter limit in the patients than in healthy controls. To ensure comparability, vessels in the controls were matched to measured vessels in the patient with RP. If, for example, all superonasal arterioles in a patient with RP were too narrow to be measured or were obscured by cataract, the superonasal arterioles were also omitted from the control measurements. Pigmentary changes in the form of bony spicules were observed in the fundi of all the patients with RP. These
Acta Ophthalmologica 2014
Results The mean oxygen saturation of retinal arterioles and venules in the patients with RP and the age- and gendermatched healthy individuals is presented in Table 1. The mean oxygen saturation in the venules of patients with RP was found to be significantly higher than in the matched normal volunteers (paired t-test, p = 0.017). No significant difference was found between the mean oxygen saturation of arterioles of patients with RP and the healthy individuals (p = 0.65). As shown in Fig. 2, the oxygen saturation in retinal venules was consistently higher in the patients with RP than the matched healthy individuals, with the exception of two cases. The mean arteriovenous (AV) difference tended to be lower in patients with RP compared with normal individuals, as indicated in Table 1, although the difference between the groups was not significant (p = 0.27). The mean diameter of the arterioles and venules of the patients with RP was found to be significantly narrower than in the normal individuals (p < 0.0001), as shown in Table 1. Representative images from one patient and a healthy age- and gender-matched volunteer are shown in Fig. 3. Cataract and narrow vessel diameter in patients with RP resulted in reduced number of measurable vessels. As described in the Methods section, vessels in the control eyes were
Table 1. Oxygen saturation and vessel diameter*.
Oxygen saturation (%) Arterioles Venules AV difference Diameter (pixels) Arterioles Venules
Healthy (n)
RP (n)
Paired t-test
90.9 1.2 (21) 53.4 4.8 (28) 36.1 4.2 (21)
91.7 3.7 (7) 58.0 6.2 (10) 32.8 5.4 (7)
p = 0.65 p = 0.017 p = 0.27
11.4 1.2 (21) 15.3 1.7 (28)
8.9 1.6 (7) 10.1 1.2 (10)
p < 0.0001 p < 0.0001
* The table shows means and standard deviations. For the healthy individuals, a mean was first calculated for the controls for each patient with RP. The table shows the mean of those means.
Arterioles
100
Oxygen saturation (%)
Oxygen saturation (%)
changes were not likely to affect the measurements in a significant way and unlikely to invalidate the assumptions of the model. Other minor pigmentary changes in the fundi may have had corresponding effects on the oxygen saturation measures, but as these changes were likely to be only minor, they were not taken into account. Differences in vessel diameter are known to cause artificial differences in measured oxygen saturation (Beach et al. 1999; Hammer et al. 2008). An attempt was made to correct for this using the method described by Geirsdottir et al. (2012). The study was approved by the National Bioethics Committee of Iceland and the Icelandic Data Protection Authority and adhered to the tenets of the Declaration of Helsinki. All subjects provided informed consent before participating in the study.
90
80
70
Healthy
Retinitis pigmentosa
Venules
70
60
50
40
Healthy
Retinitis pigmentosa
Fig. 2. Retinal vessel oxygen saturation in arterioles (left) and venules (right). Each line between data points connects a patient with RP and an average for 1–3 age- and gender-matched healthy individuals.
matched to the measured vessels in patients with RP. In three of the patients, no arterioles with diameter above 6 pixels (approximately 56 µm) were detected reliably. A post hoc power analysis was performed, using the observed standard deviations and a significance level of p = 0.05. The results were that for arterioles, the calculated power to detect a difference between patients with RP and healthy individuals is below 90% if the real difference is
Retinal vessel oxygen saturation and vessel diameter in retinitis pigmentosa Thor Eysteinsson,1,2 Sveinn H. Hardarson,2 David Bragason2 and Einar Stefansson2 1
Physiology, University of Iceland, Reykjavik, Iceland Ophthalmology, University of Iceland/Landspitali University Hospital, Reykjavik, Iceland
2
ABSTRACT. Purpose: To assess retinal vessel oxygen saturation and retinal vessel diameter in retinitis pigmentosa. Methods: A retinal oximeter (Oxymap ehf., Reykjavik, Iceland) was used to measure retinal vessel oxygen saturation and vessel diameter in ten patients with retinitis pigmentosa (RP) (mean age 49 years, range 23–71 years). Results were compared with age- and gender-matched healthy individuals. All patients had advanced stage of the disease with visual fields restricted to the macular region. Results: Oxygen saturation in retinal venules was 58.0 6.2% in patients with RP and 53.4 4.8% in healthy subjects (p = 0.017). Oxygen saturation in retinal arterioles was not significantly different between groups (p = 0.65). The mean diameter of retinal arterioles was 8.9 1.6 pixels in patients with RP and 11.4 1.2 in healthy controls (p < 0.0001). The corresponding diameters for venules were 10.1 1.2 (RP) and 15.3 1.7 (healthy, p < 0.0001). Conclusions: Increased venous saturation and decreased retinal vessel diameter suggest decreased oxygen delivery from the retinal circulation in retinitis pigmentosa. This is probably secondary to tissue atrophy and reduced oxygen consumption. Key words: oxygen – retinal oximetry – retinitis pigmentosa – vessel diameter
Acta Ophthalmol. 2014: 92: 449–453 ª 2014 Acta Ophthalmologica Scandinavica Foundation. Published by John Wiley & Sons Ltd
doi: 10.1111/aos.12359
Introduction Retinitis pigmentosa (RP) is a hereditary disease characterized by slow, progressive degeneration of photoreceptors, which leads to night blindness, constriction of visual fields and eventually reduction in visual acuity (Hartong et al. 2006; Ferrari et al. 2011). A large number of mutations in several of the genes expressed in retinal photoreceptors and retinal pigment epithelial cells have been shown to cause retinitis pigmentosa (Ferrari et al. 2011). Histopathologic examinations indicate that all layers of the retina are affected, with the most profound changes in the
photoreceptor outer segments, outer nuclear layer, less changes in the ganglion cell layer and probably the least in the inner nuclear layer (Santos et al. 1997; Milam et al. 1998; Humayun et al. 1999). Choroidal and retinal blood flow has been found to decrease in RP. Pulsatile ocular blood flow is reduced and this most likely reflects decreased choroidal blood flow (Langham & Kramer 1990; Schmidt et al. 2001). More recent studies have found decreased blood velocity in the ophthalmic artery and short posterior ciliary arteries (Cellini et al. 2010) and decreased subfoveal choroidal blood flow (Falsini et al. 2011).
Additionally, Zhang et al. (2013) found decreased blood flow in a combined measurement of the retina and choroid with an MRI technique; most of the decrease probably reflecting decreased choroidal blood flow. Histopathologic changes have been seen in the eyes of patients with RP, in which the choroidal capillaries are missing in areas corresponding to regions that have lost photoreceptors, and where bony spicules have formed (Henkind & Gartner 1983; Li et al. 1995; Milam et al. 1998). Recent studies with OCT have shown decreased thickness of the choroid in RP (Adhi et al. 2013; Ayton et al. 2013; Dhoot et al. 2013). Grunwald et al. (1996) measured the retinal circulation with a laser Doppler technique and found decreased volumetric blood flow in the major retinal venules in RP. Additionally, blood velocity has been found to be lower in retinal vessels (Beutelspacher et al. 2011) and the central retinal artery (Akyol et al. 1995). In theory, the blood flow changes could be either primary and causative or secondary, an adjustment to the decreased demand for blood and oxygen due to retinal atrophy and decreased function. Oxygen supply to the outer retina (photoreceptors) derives primarily from the choroid, while the inner retina is mostly supplied by the retinal vasculature (Linsenmeier & Padnick-Silver 2000). The boundaries between these two sources of oxygen can shift whether oxygen supply or demand changes in either the inner or outer retina. Reduced oxygen consumption in the photoreceptor layer causes oxygen from the choroid to
449
Acta Ophthalmologica 2014
reach the inner retina by diffusion, reducing the need for oxygen delivery from the retinal circulation. This could result in reduced oxygen delivery by the retinal circulation and would be seen in reduced arteriovenous difference in oxygen saturation and/or reduced blood flow and vessel diameters. Another process that may be involved is that the loss of photoreceptors that occurs in RP leads to reduced neuronal activity and metabolism in inner retinal neurones, and this affects oxygen delivery by the inner retinal vasculature. In this study, we use a non-invasive retinal oximetry technique (Hardarson et al. 2006; Geirsdottir et al. 2012) to measure oxygen saturation and diameter in retinal vessels in patients with RP. The results of this study were presented previously in an abstract form (Eysteinsson et al. 2012).
570 nm
600 nm
Methods Retinal vessel oxygen saturation and vessel diameter were measured in 10 patients with RP (mean age 49 years, range 23–71 years) and compared with 10 age and gender-matched healthy individuals. A diagnosis of RP was based on the clinical findings, which included typical funduscopic appearance, compromised peripheral and night vision, and severely reduced or no detectable scotopic and photopic aand b-waves of the electroretinogram. Visual fields as determined by standard Goldmann perimetry (II and V, 4e; Octopus 101, Haag-Streit AG, Koneitz, Switzerland) were in most cases severely restricted. Subjects with other conditions affecting the retina, diabetes, severe cardiovascular or respiratory disease were excluded from the study. The dual wavelength retinal oximeter (Oxymap ehf.) consists of two digital cameras (Insight IN1800, Diagnostic Instruments Inc., Sterling Heights, MI), a custom-made optical adapter, a beam splitter and two narrow band-pass filters, coupled to a fundus camera (Topcon TRC-50DX). The oximeter captures two fundus images simultaneously, at 570 and 600 nm wavelengths, as shown in Fig. 1. Specifically designed software (OXYMAP Analyzer software 2.2.1, version 3847, Oxymap ehf.) automatically selects measurement points on the images and measures brightness on
450
Fig. 1. Retinal oximetry from a patient with RP. The images in the upper panel were obtained via 570 nm (left) and 600 nm (right) filters simultaneously and were used for calculation of ODR. A colour-coded map of haemoglobin oxygen saturation is shown in the lower panel.
the measured vessels and to the side of the vessels at each wavelength. The ratio of light absorbance at the two wavelengths is approximately linearly related to oxygen saturation (Hardarson et al. 2006; Geirsdottir et al. 2012). The pupils of the subjects were dilated with 1% tropicamide (Mydriacyl; S.A. Alcon-Couvreur N.V., Puurs, Belgium), which was in some cases supplemented with 10% phenylephrine hydrochloride (AK-Dilate, Akorn Inc., Lake Forest, IL). Images were taken in a dark room, with the only lights coming from the fundus camera and computer screen. The lowest useable setting of the aiming light was used, and the flash intensity setting was set at 50 Ws. A band-pass filter with 80 nm full width at half maximum transmittance and 585 nm centre wavelength was used to limit unnecessary light exposure from aiming light or flash. The small aperture setting of the fundus camera was used. Small pupil setting was used only if needed. Images were first taken of the right eye and then the left eye. In general, five images were taken of each eye, most with different alignments. Images used for analysis were in general number two or five in the sequence of images, sometimes later (if more images had to be
taken to achieve good quality). Time between images was approximately 30 seconds. The eye with the better quality fundus image and more measureable vessels was chosen for analysis for each patient. The analysed images had the optic disc centred in the image. All major arterioles and venules, above 6 pixels diameter (approximately 56 µm), were measured, if possible. Measurements were made from close to the optic disc and up to the next vessel branching (50–200 pixels length limit). Each patient with RP was matched to one or an average of two or three healthy controls of the same age and gender. In some patients, a part of the vasculature was obscured by cataract. Fewer vessels were above the diameter limit in the patients than in healthy controls. To ensure comparability, vessels in the controls were matched to measured vessels in the patient with RP. If, for example, all superonasal arterioles in a patient with RP were too narrow to be measured or were obscured by cataract, the superonasal arterioles were also omitted from the control measurements. Pigmentary changes in the form of bony spicules were observed in the fundi of all the patients with RP. These
Acta Ophthalmologica 2014
Results The mean oxygen saturation of retinal arterioles and venules in the patients with RP and the age- and gendermatched healthy individuals is presented in Table 1. The mean oxygen saturation in the venules of patients with RP was found to be significantly higher than in the matched normal volunteers (paired t-test, p = 0.017). No significant difference was found between the mean oxygen saturation of arterioles of patients with RP and the healthy individuals (p = 0.65). As shown in Fig. 2, the oxygen saturation in retinal venules was consistently higher in the patients with RP than the matched healthy individuals, with the exception of two cases. The mean arteriovenous (AV) difference tended to be lower in patients with RP compared with normal individuals, as indicated in Table 1, although the difference between the groups was not significant (p = 0.27). The mean diameter of the arterioles and venules of the patients with RP was found to be significantly narrower than in the normal individuals (p < 0.0001), as shown in Table 1. Representative images from one patient and a healthy age- and gender-matched volunteer are shown in Fig. 3. Cataract and narrow vessel diameter in patients with RP resulted in reduced number of measurable vessels. As described in the Methods section, vessels in the control eyes were
Table 1. Oxygen saturation and vessel diameter*.
Oxygen saturation (%) Arterioles Venules AV difference Diameter (pixels) Arterioles Venules
Healthy (n)
RP (n)
Paired t-test
90.9 1.2 (21) 53.4 4.8 (28) 36.1 4.2 (21)
91.7 3.7 (7) 58.0 6.2 (10) 32.8 5.4 (7)
p = 0.65 p = 0.017 p = 0.27
11.4 1.2 (21) 15.3 1.7 (28)
8.9 1.6 (7) 10.1 1.2 (10)
p < 0.0001 p < 0.0001
* The table shows means and standard deviations. For the healthy individuals, a mean was first calculated for the controls for each patient with RP. The table shows the mean of those means.
Arterioles
100
Oxygen saturation (%)
Oxygen saturation (%)
changes were not likely to affect the measurements in a significant way and unlikely to invalidate the assumptions of the model. Other minor pigmentary changes in the fundi may have had corresponding effects on the oxygen saturation measures, but as these changes were likely to be only minor, they were not taken into account. Differences in vessel diameter are known to cause artificial differences in measured oxygen saturation (Beach et al. 1999; Hammer et al. 2008). An attempt was made to correct for this using the method described by Geirsdottir et al. (2012). The study was approved by the National Bioethics Committee of Iceland and the Icelandic Data Protection Authority and adhered to the tenets of the Declaration of Helsinki. All subjects provided informed consent before participating in the study.
90
80
70
Healthy
Retinitis pigmentosa
Venules
70
60
50
40
Healthy
Retinitis pigmentosa
Fig. 2. Retinal vessel oxygen saturation in arterioles (left) and venules (right). Each line between data points connects a patient with RP and an average for 1–3 age- and gender-matched healthy individuals.
matched to the measured vessels in patients with RP. In three of the patients, no arterioles with diameter above 6 pixels (approximately 56 µm) were detected reliably. A post hoc power analysis was performed, using the observed standard deviations and a significance level of p = 0.05. The results were that for arterioles, the calculated power to detect a difference between patients with RP and healthy individuals is below 90% if the real difference is
